Speakers, Buzzers, and Microphones

Every phone call uses two small parts that work in opposite directions. A microphone turns sound into an electrical signal, and a speaker turns an electrical signal back into sound. A buzzer is a simpler cousin of the speaker, made to beep rather than to play music or speech. This article explains what sound is, how each part works, how to read its numbers, how to wire it up, and how to use it safely.

1. A short recap

• A current in a coil of wire makes a magnetic field, and a changing magnetic field induces a voltage in a coil (recap).
• A capacitor is two conductors separated by a gap. Its capacitance grows as the gap shrinks, and it blocks steady DC while passing AC (recap, coupling).
• Alternating current (AC) swings back and forth. The number of complete swings each second is its frequency, measured in hertz (Hz) (recap).
• Impedance is the total opposition a part offers to AC, in ohms (Ω). A coil's opposition rises with frequency (recap).
• A transistor lets a small signal switch or control a much larger current (recap).
• A transducer is any device that converts energy from one form to another; every sensor is one (recap).

2. Sound, in electrical terms

Sound is a pressure wave. A vibrating object pushes on the air next to it, squeezing the air molecules together (a compression) and then pulling back and leaving them spread out (a rarefaction). That pattern travels outward at the speed of sound, 343 meters per second in air at 20 °C.^[1]

Three numbers describe a steady tone:

• Frequency, in hertz: how many compressions pass each second. We hear frequency as pitch.^[1] Normal human hearing runs from about 20 to 20,000 Hz (20 kHz).^[2] The top of that range falls with age.^[3]
• Wavelength: the distance from one compression to the next. Speed = frequency × wavelength,^[1] so a 20 Hz bass note is about 17 m long and a 20 kHz whistle about 1.7 cm (calculated). Low notes need big speakers to push enough air; that is why a large woofer handles low frequencies and a small tweeter handles high ones.^[1]
• Amplitude: how far the pressure swings above and below normal air pressure, in pascals (Pa).

Decibels

The ear handles an enormous range: a sound that damages hearing in seconds is about 10¹² (a million million) times as intense as the faintest one we can hear.^[4] So loudness is measured on a logarithmic scale in decibels (dB). Sound pressure level (SPL) compares a pressure p with 20 micropascals (µPa), roughly the quietest sound a person can hear:^[5]

SPL = 20 × log₁₀(p / 20 µPa)  dB

A pressure of 1 Pa works out to 94 dB SPL (calculated); microphone makers use it as their standard test level. Some useful rules: doubling the pressure adds 6 dB,^[5] ten times the sound power adds 10 dB,^[4] and doubling your distance from a small source such as a buzzer takes away about 6 dB.^[6] The ear is most sensitive between about 2,000 and 5,000 Hz,^[2] so sound meters often apply A-weighting, which discounts frequencies the ear hears poorly; such readings are written dBA.^[3] Figure 2 puts everyday sounds on the scale.

3. Transducers for sound

An audio transducer converts between electrical signals and sound. Most rely on one of three physical effects, and each one works in both directions:

• Magnetic force and induction. A current-carrying wire in a magnetic field feels a force, so a signal current can push a coil and the cone attached to it.^[7] Run the process backward: moving a coil through a magnetic field induces a voltage in it.^[8] Speakers, magnetic buzzers and dynamic microphones work this way.
• Changing capacitance. If a capacitor holds a fixed charge, moving one plate changes its capacitance and so its voltage. Condenser, electret and MEMS microphones work this way.^[9]​^[10]
• The piezoelectric effect. Some crystals and ceramics bend when a voltage is applied across them, and make a voltage when they are squeezed.^[6] Piezo buzzers work this way. (The same effect returns in Crystals and Oscillators.)

4. Speakers

How a speaker is built

The most common speaker is the moving-coil or dynamic speaker. Its main parts are:^[11]

• a voice coil, a light coil of fine wire wound on a thin tube;
• a strong permanent magnet with steel parts (the plate and yoke) that concentrate its field across a narrow ring-shaped gap, where the voice coil sits;
• a cone or diaphragm, glued to the voice coil, that pushes on the air;
• a flexible suspension that holds the cone centered and lets it move in and out, attached to a rigid frame.^[12]

How it works

A magnetic field pushes sideways on a current-carrying wire. For a straight wire at right angles to the field, the force is:^[7]

F = B × I × l

where B is the field in teslas, I the current in amperes and l the length of wire in the field in meters. In a speaker the field crosses the gap from the center outward, so every turn of the voice coil is at right angles to it, and all the small forces add up to a push along the axis. Reverse the current and the push reverses. A music signal is an alternating current, so the cone moves in and out in step with it and makes sound.^[11]

Speaker datasheets combine field and wire length into one number, the BL product, in tesla-meters (T·m). The Dayton Audio PC83-8, a 3-inch speaker, has BL = 3.5 T·m,^[13] so each ampere of current pushes with 3.5 newtons. At about 0.35 A, the current of a 1-watt signal in an 8 Ω speaker, the force is about 1.2 N (calculated).

The cone and its suspension also have a natural rate of vibration, the resonant frequency, set by the mass of the moving parts and the stiffness of the suspension: lighter, stiffer parts resonate higher.^[12] This is the mechanical version of the coil-and-capacitor resonance in RCL circuits (recap). The resonant frequency gives a rough guide to how deep a speaker's bass can go.^[11]

The same force that moves a voice coil also turns electric motors, and the process runs backward too: wiggle the cone of a speaker by hand and its coil, moving through the magnetic field, makes a small voltage (this follows from the law of induction).^[8]

Units and markings

• Impedance, in ohms, sets how much current the speaker draws from an amplifier: the lower the impedance, the more power it draws.^[11] The printed figure is the nominal impedance, a round number such as 4, 8 or 16 Ω. The real value changes with frequency (Figure 5). The PC83-8 is rated 8 Ω, but its coil's plain DC resistance, called R_e, is 6.8 Ω, and its coil inductance is 0.38 mH.^[13]
• Power handling, in watts. The nominal or rated power is what the speaker can take continuously; the maximum power is what it can survive for a short time.^[11] The CSS-50508N is rated 0.25 W nominal and 0.5 W maximum.^[14] The PC83-8 handles 30 W.^[13]
• Sensitivity, in dB SPL: how loud the speaker plays for a standard input, usually 1 W (2.83 V into 8 Ω), measured 1 m in front.^[5] The PC83-8 gives 83.8 dB at 2.83 V and 1 m.^[13] Small speakers are often measured differently: the CSS-50508N gives 85 dB at 0.25 W and 0.5 m.^[14] Same Sky's correction rules (add 10 log₁₀ of the power ratio and 20 log₁₀ of the distance ratio) convert that to about 85 dB at 1 W and 1 m, so the two are close (calculated).^[5]
• Frequency response: a graph of output level against frequency. No speaker is perfectly flat.^[11] The PC83-8 is usable from 80 to 20,000 Hz.^[13] The 50 mm CSS-50508N resonates at 380 Hz and is flat only from about 800 Hz to 3 kHz; small speakers do not reproduce low frequencies well.^[15]
• Markings: often just the impedance and power stamped on the magnet or frame, such as "8Ω 0.25W" on the CSS-50508N.^[14]

Common types

• Full-range drivers such as the PC83-8 try to cover most of the audible range with one cone.^[13]
• Woofers and tweeters split the job: a large cone for low notes, a small light one for high notes.^[1]​^[12]
• Miniature speakers, a few centimeters across or less, fit phones and gadgets. Digital processing and a carefully designed enclosure help their weak bass.^[11]
• By magnet: ferrite (ceramic) magnets are cheap and heavy; neodymium (NdFeB) magnets are the strongest, so they suit small, loud speakers, but they can shatter.^[11]

5. Driving a speaker

Never connect a speaker directly to a microcontroller pin. A 5 V output across a 6.8 Ω coil would try to drive over 700 mA (calculated), and the ATmega328P of a classic Arduino Uno is rated at an absolute maximum of 40 mA per pin.^[16] The signal needs an amplifier, a circuit that makes a larger copy of it using power from a supply.

A first amplifier: the LM386

The LM386 is a classic small audio amplifier chip, an op-amp-style integrated circuit (recap). It runs from 4 to 12 V, drives speakers from 4 to 32 Ω, and delivers about 325 mW into 8 Ω from a 6 V supply.^[17] Its voltage gain, how many times bigger the output is than the input, is 20 with pins 1 and 8 left open, and 200 with a 10 µF capacitor between them.^[17]

The output sits at half the supply voltage when there is no signal,^[17] so a large capacitor (250 µF in TI's circuit) connects it to the speaker. The capacitor passes the audio and blocks the DC, so no steady current flows through the coil (coupling). With an 8 Ω speaker it begins to cut frequencies below about 80 Hz (calculated, 1/(2πRC)). TI's minimal circuit also places a 10 Ω resistor in series with a 0.05 µF capacitor from the output to ground; copy it as shown.^[17]

Impedance matching

In audio, "matching" mostly means keeping each part inside the range the other expects:

• Speaker to amplifier. An amplifier's output voltage is limited, and power is P = V²/R, so halving the impedance roughly doubles the current and the power drawn. Use speakers whose impedance is at or above the amplifier's minimum, or you risk damaging the equipment.^[11] Two 8 Ω speakers in parallel make 4 Ω; in series, 16 Ω (recap).
• Microphone to amplifier. Here the aim is to keep the signal voltage. A source with internal impedance feeding an input forms a voltage divider (recap). An electret microphone with a 2.2 kΩ output impedance^[18] feeding the LM386's 50 kΩ input^[17] keeps about 96% of its voltage (calculated). So a microphone input should have a much higher impedance than the microphone.
• Transformers can convert one impedance to another by their turns ratio (recap).

6. Buzzers

A buzzer makes a loud tone in a narrow range of frequencies; it does not need to reproduce speech or music.^[15] Buzzers come in two technologies, magnetic and piezo, and two configurations:^[6]

• An indicator has its drive circuit built in. Apply the rated DC voltage and it sounds at a fixed frequency.
• A transducer has no drive circuit. You must feed it a square wave, a signal that flips between on and off at the frequency you want. That takes more work but allows tones, sirens and chimes.

Magnetic buzzers

In a magnetic buzzer a current through a coil makes a magnetic field that pulls a flexible disk of ferromagnetic metal (one that magnets attract, such as iron). When the current stops, the disk springs back. Switching the current on and off makes the disk vibrate, like a tiny speaker cone.^[6] It is current-driven: the coil's impedance and the applied voltage set the current. Magnetic buzzers typically run on 1.5 to 12 V and draw more than 20 mA.^[6]

Example: Same Sky's CEM-1206S magnetic transducer is 12 mm across, rated 5 V peak with a 2,400 Hz square wave, draws 45 mA, has a coil resistance of 47 Ω, and produces at least 85 dBA at 10 cm.^[19] Held on steady DC, the same coil would draw about 106 mA (calculated, 5 V ÷ 47 Ω) and make no sound beyond a click.

Piezo buzzers

A piezo buzzer is a thin disk of piezoelectric ceramic with metal contacts on both faces, held at its edges inside a case. A voltage across the disk makes it flex, and the flexing makes sound.^[6] Electrically, a piezo buzzer behaves like a capacitor, not a coil, and it is voltage-driven: its loudness follows the drive voltage fairly steadily.^[6]​^[20] Piezo buzzers typically use 12 to 220 V and less than 20 mA, and can reach higher sound levels than magnetic buzzers, though they need more space to do so.^[6]

Example: Same Sky's CPT-2475-92T piezo transducer is rated 10 V peak-to-peak (operating from 1 to 30 V), resonates at 4,000 Hz, draws 8 mA, gives 92 dB at 10 cm, and has a capacitance of 21 nF.^[21] At 4 kHz that capacitance has a reactance of about 1.9 kΩ (calculated; reactance). Peak-to-peak (V_p-p) is the full swing from the lowest to the highest voltage, so 10 V_p-p means 5 V each side of zero.^[20]

Every buzzer is loudest at its resonant frequency, so drive it there unless you need other tones.^[6]​^[12]

Sources: Same Sky buzzer guide and data sheets.^[6]​^[19]​^[21]

Driving a buzzer

A transducer needs a square wave, which a microcontroller pin or a 555 timer can produce (555 oscillator), but usually through a transistor switch (recap):^[6]

• Magnetic transducer: a BJT or MOSFET switches the coil. A diode across the coil is required to clamp the voltage spike, the flyback, that a coil produces when its current is switched off quickly (flyback diode). The CEM-1206S's 45 mA is already over the ATmega328P's 40 mA limit, so the transistor is not optional.^[19]​^[16]
• Piezo transducer: no flyback diode is needed, because a piezo has little inductance, but a resistor across it is needed to discharge it when the switch opens.^[6]
• Louder piezos: a full-bridge driver of four switches applies the supply first one way, then the other, doubling the peak-to-peak voltage and adding about 6 dB.^[6] Such bridge circuits are also used to drive electric motors.^[20] A resonant driver, an inductor paired with the piezo's own capacitance, can build up a voltage many times the supply.^[20]

7. Microphones

Units and markings

• Sensitivity: how much output voltage the microphone gives for a standard sound of 1 Pa (94 dB SPL) at 1 kHz. It is given in decibels relative to 1 volt per pascal, dBV/Pa: S = 20 log₁₀(V / 1 V). Almost all microphones give far less than 1 V, so the number is negative; a microphone with twice the output is 6 dB more sensitive.^[5] The CUI CMA-4544PF-W electret microphone is −44 ±2 dB,^[18] which is 6.3 mV per pascal (calculated, 10^−44/20 V).
• Signal-to-noise ratio (SNR): how far the 1 Pa signal stands above the microphone's own hiss, in dB; 60 dBA for the CMA-4544PF-W.^[18]
• Frequency response: 20 to 20,000 Hz for that part.^[18] Some microphones are deliberately shaped, for example with a peak at 2–10 kHz to make voices clearer.^[22]
• Directionality (polar pattern): omnidirectional microphones hear all directions equally; cardioid (heart-shaped) ones mainly hear what is in front of them.^[22]

Dynamic microphones

A moving-coil dynamic microphone is a speaker run backward: a diaphragm, a voice coil and a magnet form a "sound-driven electrical generator". Sound moves the diaphragm, the coil moves in the magnetic field, and a voltage is induced that follows the sound wave.^[22] That is Faraday's law: a changing magnetic flux through a coil induces a voltage.^[8] Dynamic microphones need no power, are rugged and handle very loud sounds without distorting, so they are common on stage.^[22] A ribbon microphone is a variant that uses a thin strip of metal between magnetic poles; it is delicate, and its output is low.^[22]

Condenser and electret microphones

A condenser microphone ("condenser" is an old word for capacitor) uses a thin conductive diaphragm and a charged backplate as the two plates of a capacitor. Sound moves the diaphragm and changes the gap.^[22] Since C = Q/V, a fixed charge Q on a changing capacitance C gives a changing voltage V.^[9] To use this signal, every condenser microphone contains active electronics, a preamp, so it needs power, either from a battery or from a supply called phantom power.^[22]

An electret is a material with a permanent built-in electric charge, much as a magnet has a permanent magnetic field. Electret condenser microphones (ECMs) put it on the diaphragm, so they need no voltage to charge the plates.^[9] The small voltage changes are buffered by a JFET, a type of transistor, inside the capsule (recap). Outside, you add a load resistor from the supply and a DC-blocking capacitor.^[10] The CMA-4544PF-W, for example, is a metal can 9.7 mm across.^[18]

MEMS microphones

A MEMS (micro-electro-mechanical system) microphone uses the same capacitor principle, but its diaphragm and backplate are etched into a silicon chip. A second chip in the same package amplifies the signal, and in digital types also converts it to numbers with an analog-to-digital converter (ADC). A small hole, the port, lets sound in through the lid or through the circuit board underneath.^[10]

MEMS microphones come with three kinds of output:^[23]​^[10]

• Analog: a voltage, already amplified, with a fairly low output impedance. It uses less power than digital types.
• PDM (pulse density modulation): a stream of single bits on a data line, timed by a clock line, where the share of 1s follows the sound. Two microphones can share the same two wires for stereo.
• I²S: a standard digital audio format with ready-made samples, which can connect straight to a microcontroller with no separate converter.

Example: Knowles' SPH0645LM4H-B is a bottom-port I²S microphone that runs from 1.62 to 3.6 V and draws about 600 µA. Its sensitivity is −26 dBFS,^[24] meaning a 94 dB sound produces a signal 26 dB below the largest number the output can carry (full scale).

MEMS microphones are tiny, survive machine soldering and are closely matched, which suits arrays of several microphones; electrets offer more mounting options, sealed and directional versions, and wider supply ranges.^[10]

Sources: Shure, Same Sky, CUI and Knowles.^[22]​^[10]​^[18]​^[24]

8. Microphone circuits

Figure 10 shows the usual way to connect an electret microphone. The JFET inside acts as a current-controlling valve: a resistor from the supply (2.2 kΩ, as in the CMA-4544PF-W's test circuit^[18]) turns its changing current into a changing voltage on top of a steady DC level. A capacitor passes the audio and blocks that DC level.^[10] The capacitor and the input resistance after it form a high-pass filter whose corner, 1/(2πRC), must be below 20 Hz to keep all the audio.^[23] With 1 µF into 10 kΩ, the corner is about 16 Hz (calculated; RC filters).

How much gain is needed? Speech at 1 m is about 40 to 60 dB SPL.^[3] At 60 dB, the pressure is 0.02 Pa, and a −44 dBV/Pa microphone gives about 0.13 mV (calculated). An Arduino ADC step is 4.9 mV (recap), so the raw signal is far too small to read. Adafruit's electret breakout pairs the capsule with a MAX4466 op-amp whose gain is adjustable from 25 to 125.^[25] At 125, the 0.13 mV becomes about 16 mV: still only a few ADC steps for quiet speech, but a loud 94 dB sound gives about 0.8 V (calculated).

The amplifier's output sits at half its supply when all is quiet, 1.65 V on 3.3 V, and the sound swings above and below that level, so the ADC can read both halves of the wave.^[25] To measure loudness, Adafruit's example takes readings for 50 ms (one full cycle of a 20 Hz tone) and subtracts the lowest from the highest.^[25] A MEMS microphone with I²S output skips all of this and delivers numbers directly.^[23] To start programming a board to read these signals, see the Microcontrollers pages of this site.

9. Testing

• Speakers. With everything unpowered, measure across the terminals with a multimeter. You should see the coil's DC resistance, a bit below the nominal impedance: about 6.8 Ω for the 8 Ω PC83-8.^[13] An open circuit means a broken coil. Then play a quiet tone: a scratchy, buzzing sound on low notes, which manufacturers test for as "rub and buzz", suggests a damaged or off-center voice coil.^[11]
• Magnetic buzzers. The coil resistance should match the datasheet: 40 to 54 Ω for the CEM-1206S.^[19] Check whether you have an indicator (needs DC) or a transducer (needs a square wave); a transducer on DC only clicks.^[6]
• Piezo buzzers. A resistance meter reads open circuit, because the part is a capacitor; a capacitance meter should read close to the datasheet value, about 21 nF for the CPT-2475-92T.^[21]
• Electret microphones. Connect the supply through the load resistor and measure the DC voltage at the output. With 3 V and 2.2 kΩ, at most 0.5 mA flows,^[18] so the output sits somewhere above about 1.9 V (calculated). If it reads the full supply voltage, no current is flowing: check the wiring against the datasheet. Then speak into it while watching an oscilloscope.
• MEMS microphones. A digital MEMS microphone needs its clock: the SPH0645LM4H-B drops into sleep when the clock falls below about 900 kHz, so no clock means no data.^[24]

10. Everyday uses

• Speakers are in phones, laptops and alarms.^[11] A phone needs only the range of the voice, well short of 20 Hz to 20 kHz.^[15]
• Buzzers sound in home appliances, security panels, door-entry systems and computer equipment,^[26] and give the beeps, warbles and sirens of notification and alarm products.^[15]​^[6]
• Microphones: dynamic types for stage vocals;^[22] electret types in noise-canceling headsets for call centers, helicopters and race car drivers;^[9] and MEMS types in wearables, home assistants and voice-controlled devices.^[10]​^[23]

11. Safety

• Protect your hearing. NIOSH, the U.S. workplace safety institute, treats repeated exposure at or above 85 dBA as hazardous and recommends a limit of 85 dBA averaged over 8 hours, halving the allowed time for every 3 dB louder. Hearing loss from noise is 100% preventable.^[27] A small buzzer can give 85 to 92 dB at 10 cm,^[19]​^[21] and every halving of the distance adds about 6 dB,^[6] so never test one held against your ear. Start headphones and amplifiers at low volume.
• Protect your microcontroller. Never drive a speaker or magnetic buzzer directly from a pin; use an amplifier or a transistor, and always fit a flyback diode across a coil (inductive kickback).^[16]​^[6]
• Protect your speakers. Stay within the nominal power rating for continuous use,^[11] keep DC out of the voice coil with a coupling capacitor,^[17] and never connect a load below the amplifier's minimum impedance.^[11]
• Piezo drivers can make high voltages. Bridge and resonant drivers raise the voltage across the piezo, in the resonant case to many times the supply,^[20] and piezo buzzers themselves can be rated up to 220 V.^[6] Respect the part's peak-to-peak rating.^[20]
• Phantom power and ribbon microphones. Phantom power applied to a ribbon microphone can damage it.^[22]
• Mains-powered amplifiers contain lethal voltages. Build with batteries or low-voltage adapters (recap). MEMS microphones are chips and need the usual static precautions (handling chips).
• Magnets from speakers. Speaker magnets, especially neodymium ones, are strong. Keep loose magnets away from children: two or more swallowed magnets can attract each other inside the body and cause serious injury, so seek medical help immediately if you suspect it.^[28]

The whole story in brief

• Sound is a pressure wave; frequency is pitch (about 20 Hz to 20 kHz for people), and loudness is measured in dB SPL against 20 µPa.
• Audio transducers use magnetic force and induction, changing capacitance, or the piezoelectric effect, and each effect works in both directions.
• A dynamic speaker is a voice coil in a magnet's gap, pushing a cone with force F = BIl; its impedance, power handling, sensitivity and frequency response describe it.
• Speakers need an amplifier; keep speaker impedance at or above the amplifier's minimum, and give microphones a high-impedance input.
• Magnetic buzzers are current-driven coils that need a flyback diode; piezo buzzers are voltage-driven capacitors that need a reset resistor; indicators need DC, transducers a square wave.
• Dynamic microphones generate their own voltage; condenser, electret and MEMS microphones read a changing capacitance and need power. Sensitivity in dBV/Pa says how big the signal is.
• Protect your ears: 85 dBA is where repeated exposure becomes hazardous.

Coming up in this section: 20 Motors.

Sources

All links were checked in October 2026. External links open in a new tab.

1. OpenStax, College Physics 2e, section 17.2, "Speed of Sound, Frequency, and Wavelength": https://openstax.org/books/college-physics-2e/pages/17-2-speed-of-sound-frequency-and-wavelength
2. OpenStax, College Physics 2e, section 17.6, "Hearing": https://openstax.org/books/college-physics-2e/pages/17-6-hearing
3. Same Sky (formerly CUI Devices), "The Basics of Sound Pressure Level and Decibels" (blog): https://www.sameskydevices.com/blog/the-basics-of-sound-pressure-level-and-decibels
4. OpenStax, College Physics 2e, section 17.3, "Sound Intensity and Sound Level": https://openstax.org/books/college-physics-2e/pages/17-3-sound-intensity-and-sound-level
5. Same Sky, "Understanding and Measuring Sensitivity in Audio Components" (blog): https://www.sameskydevices.com/blog/understanding-and-measuring-sensitivity-in-audio-components
6. Nick Grillone, Same Sky, "Buzzer Basics: Technologies, Tones, and Drive Circuits" (blog): https://www.sameskydevices.com/blog/buzzer-basics-technologies-tones-and-driving-circuits
7. OpenStax, College Physics 2e, section 22.7, "Magnetic Force on a Current-Carrying Conductor": https://openstax.org/books/college-physics-2e/pages/22-7-magnetic-force-on-a-current-carrying-conductor
8. OpenStax, College Physics 2e, section 23.2, "Faraday's Law of Induction: Lenz's Law": https://openstax.org/books/college-physics-2e/pages/23-2-faradays-law-of-induction-lenzs-law
9. Same Sky, "All You Need to Know About Electret Condenser Microphones" (blog): https://www.sameskydevices.com/blog/all-you-need-to-know-about-electret-condenser-microphones
10. Nick Grillone, Same Sky, "Comparing MEMS and Electret Condenser (ECM) Microphones" (blog): https://www.sameskydevices.com/blog/comparing-mems-and-electret-condenser-microphones
11. Nick Grillone, Same Sky, "All You Need to Know About Speakers in Audio Design" (blog): https://www.sameskydevices.com/blog/all-you-need-to-know-about-speakers-in-audio-design
12. Same Sky, "Understanding the Impact of Resonance and Resonant Frequency in Audio Design" (blog): https://www.sameskydevices.com/blog/understanding-the-impact-of-resonance-and-resonant-frequency-in-audio-design
13. Dayton Audio, PC83-8 3" Full-Range Poly Cone Driver specification sheet (PDF): https://www.daytonaudio.com/images/resources/295-156--dayton-audio-pc83-8-specifications.pdf
14. Same Sky, CSS-50508N speaker data sheet (rev. 1.04, February 2025, PDF): https://www.sameskydevices.com/product/resource/css-50508n.pdf
15. Same Sky, "Understanding Audio Frequency Range in Audio Design" (blog): https://www.sameskydevices.com/blog/understanding-audio-frequency-range-in-audio-design
16. Microchip Technology, ATmega48A/PA/88A/PA/168A/PA/328/P data sheet (DS40002061B, 2020), sections 29.1 and 30.1, "Absolute Maximum Ratings" (PDF): https://ww1.microchip.com/downloads/aemDocuments/documents/MCU08/ProductDocuments/DataSheets/ATmega48A-PA-88A-PA-168A-PA-328-P-DS-DS40002061B.pdf
17. Texas Instruments, LM386 Low Voltage Audio Power Amplifier data sheet (SNAS545D, revised August 2023, PDF): https://www.ti.com/lit/ds/symlink/lm386.pdf
18. CUI Devices (now Same Sky), CMA-4544PF-W electret condenser microphone data sheet (PDF): https://www.sameskydevices.com/product/resource/cma-4544pf-w.pdf
19. Same Sky, CEM-1206S magnetic buzzer transducer data sheet (February 2025, PDF): https://www.sameskydevices.com/product/resource/cem-1206s.pdf
20. Same Sky, "How to Increase the Audio Output of a Piezoelectric Transducer Buzzer" (blog): https://www.sameskydevices.com/blog/how-to-increase-the-audio-output-of-a-piezoelectric-transducer-buzzer
21. Same Sky, CPT-2475-92T piezo buzzer transducer data sheet (September 2024, PDF): https://www.sameskydevices.com/product/resource/cpt-2475-92t.pdf
22. Shure, "Microphone Basics: Transducers, Polar Patterns, & Frequency Response" (March 2015): https://www.shure.com/en-us/insights/microphone-basics-transducers-polar-patterns-frequency-response
23. Nick Grillone, Same Sky, "Analog or Digital: How to Choose the Right MEMS Microphone Interface" (blog): https://www.sameskydevices.com/blog/analog-or-digital-how-to-choose-the-right-mems-microphone-interface
24. Knowles, SPH0645LM4H-B I²S output digital microphone data sheet (rev. B, PDF, hosted by Adafruit): https://cdn-shop.adafruit.com/product-files/3421/i2S+Datasheet.PDF
25. Bill Earl, Adafruit Learning System, "Adafruit Microphone Amplifier Breakout" (updated June 2024, PDF): https://cdn-learn.adafruit.com/downloads/pdf/adafruit-microphone-amplifier-breakout.pdf
26. Bruce Rose (CUI Inc), "What you need to know about buzzers: reviewing the basic types and technologies", Power Electronic Tips (March 2019): https://www.powerelectronictips.com/basic-types-technologies-buzzers/
27. U.S. Centers for Disease Control and Prevention, NIOSH, "Understand Noise Exposure" (January 2024): https://www.cdc.gov/niosh/noise/prevent/understand.html
28. U.S. Consumer Product Safety Commission, "Magnets" safety education center: https://www.cpsc.gov/Safety-Education/Safety-Education-Centers/Magnets